Tag
This paper introduces SCOPE-FL, a hierarchical federated learning framework that uses the Top Trading Cycle algorithm to ensure strategy-proofness and Pareto efficiency in client selection, with reward distribution via Shapley value approximation and blockchain-based execution.