An Incentive Mechanism for Federated Learning in Wireless Cellular network An Auction Approach
Abstract : Federated Learning (FL) is a distributed learning framework that can deal with the distributed issue in machine learning and still guarantee high learning performance. However, it is impractical that all users will sacrifice their resources to join the FL algorithm. This motivates us to study the incentive mechanism design for FL. In this paper, we consider a FL system that involves one base station (BS) and multiple mobile users. The mobile users use their own data to train the local machine learning model, and then send the trained models to the BS, which generates the initial model, collects local models and constructs the global model. Then, we formulate the incentive mechanism between the BS and mobile users as an auction game where the BS is an auctioneer and the mobile users are the sellers. In the proposed game, each mobile user submits its bids according to the minimal energy cost that the mobile users experiences in participating in FL. To decide winners in the auction and maximize social welfare, we propose the primal-dual greedy auction mechanism. The proposed mechanism can guarantee three economic properties, namely, truthfulness, individual rationality and efficiency. Finally, numerical results are shown to demonstrate the performance effectiveness of our proposed mechanism.
? Several existing works studied the implementation problem of FL over wireless link
? Most of the existing work has focused on designing learning algorithms with provable convergence time, but other issues such as incentive mechanism are unexplored.
? An intuitive idea is to reward participants according to their contributions, following the existing incentive mechanism designs for many other scenarios.
? Although there exist some preliminaries studies, they cannot be combined with incentive mechanism design for federated learning.
? Mobile devices equipped with specialized hardware architectures and computing engines can handle the machine learning problem effectively.
? The perspective of the BS, we formulate the winner selection problem in the auction game as the social welfare maximization problem which is a NP-hard problem.
? We propose a primal-dual greedy algorithm to deal with the NP-hard problem in selecting the winning users and critical value based payment.
? The work in also considered the latency and energy consumption minimization problem for the case of asynchronous transmission.
• The proposed mechanism can provide the guarantee of efficiency while approximate truthfulness and individual rationality are ensured simultaneously.
• Federated learning has been proposed to enable distributed computing nodes to collaboratively train models without exposing their own data.
• In this paper, we propose a novel incentive mechanism design that integrates model updating using fresh data for federated learning in IoT applications.
• Due to the unique challenges of unshared information and difficulties of contribution evaluation in federated learning, we propose the DRL-based incentive mechanism to address these issues.
? The proposed mechanism can guarantee three economic properties, namely, truthfulness, individual rationality and efficiency.
? Moreover, because the wireless resource is limited, the BS needs to allocate the resources reasonably to avoid the congestion, guarantee the model training performance and optimize the total utilities of the BS and mobile users.
? In spite of the above mentioned benefits of FL, there are remaining challenges of having an efficient FL framework.
? We also proposed auction mechanism is truthful, individual rational and computational efficient.
? This proposed algorithm in can improve the training efficiency and solve the fairness issue.
|