Collaboration in Participant-Centric Federated Learning: A Game-Theoretical Perspective

Abstract

Federated learning (FL) is a promising distributed framework for collaborative artificial intelligence model training while protecting user privacy. A bootstrapping component that has attracted significant research attention is the design of incentive mechanism to stimulate user collaboration in FL. The majority of works adopt a broker-centric approach to help the central operator to attract participants and further obtain a well-trained model. Few works consider forging participant-centric collaboration among participants to pursue an FL model for their common interests, which induces dramatic differences in incentive mechanism design from the broker-centric FL. To coordinate the selfish and heterogeneous participants, we propose a novel analytic framework for incentivizing effective and efficient collaborations for participant-centric FL. Specifically, we respectively propose two novel game models for contribution-oblivious FL (COFL) and contribution-aware FL (CAFL), where the latter one implements a minimum contribution threshold mechanism. We further analyze the uniqueness and existence for Nash equilibrium of both COFL and CAFL games and design efficient algorithms to achieve equilibrium solutions. Extensive performance evaluations show that there exists free-riding phenomenon in COFL, which can be greatly alleviated through the adoption of CAFL model with the optimized minimum threshold.

Quality-Aware Incentive Mechanism Design Based on Matching Game for Hierarchical Federated Learning

“…Given a fixed K l , the upper bound of expected difference of the edge server l F (ω K l l )−F (ω * l ) is decreasing and convex function respect to the amount of data of all devices serving edge server l and satisfies diminishing marginal effect. According to [20], [24], after K l edge aggregations, the model improvement Λ l is dominated by term K l n∈S l D l n and can be approximately expressed as:…”

“…Currently, there are few studies on incentive mechanism design for HFL, with most existing works focusing on incentive mechanism design for conventional FL. [20] proposed a novel analytic framework for incentivizing effective and efficient collaborations for participant-centric FL, and designed efficient algorithms to achieve equilibrium solutions. The authors proposed the reputation-aware hedonic coalition formation game in [5] to improve the sustainable efficiency of the FL system while taking into account the incentive design for devices' marginal contributions in FL system.…”

Contribution Matching-Based Hierarchical Incentive Mechanism Design for Crowd Federated Learning

“…According to [9], heterogeneous model owners with different social statuses and contributions have different needs for model aggregation weight assignment. However, limited by the privacy-preserving nature of CFL, it is difficult for the cloud server to directly access the data to quantify the value of model updates contributed by each model owner [10]. If there is no control to ensure fairness in the aggregation weight assignment, rational model owners will lose their original market share by sharing a consistent training model, and thus choose to disengage from the cloud server.…”